Computer-aided Detection of Squamous Carcinoma of the Cervix in Whole Slide Images

Ye Tian,Li Yang,Wei Wang,Jing Zhang,Qing Tang,Mili Ji,Yang Yu,Yu Li,Hong Yang,Airong Qian

DOI: https://doi.org/10.48550/arXiv.1905.10959

2019-05-27

Abstract:Goal: Squamous cell carcinoma of cervix is one of the most prevalent cancer worldwide in females. Traditionally, the most indispensable diagnosis of cervix squamous carcinoma is histopathological assessment which is achieved under microscope by pathologist. However, human evaluation of pathology slide is highly depending on the experience of pathologist, thus big inter- and intra-observer variability exists. Digital pathology, in combination with deep learning provides an opportunity to improve the objectivity and efficiency of histopathologic slide analysis. Methods: In this study, we obtained 800 haematoxylin and eosin stained slides from 300 patients suffered from cervix squamous carcinoma. Based on information from morphological heterogeneity in the tumor and its adjacent area, we established deep learning models using popular convolution neural network architectures (inception-v3, InceptionResnet-v2 and Resnet50). Then random forest was introduced to feature extractions and slide-based classification. Results: The overall performance of our proposed models on slide-based tumor discrimination were outstanding with an AUC scores > 0.94. While, location identifications of lesions in whole slide images were mediocre (FROC scores > 0.52) duo to the extreme complexity of tumor tissues. Conclusion: For the first time, our analysis workflow highlighted a quantitative visual-based slide analysis of cervix squamous carcinoma. Significance: This study demonstrates a pathway to assist pathologist and accelerate the diagnosis of patients by utilizing new computational approaches.

Computer Vision and Pattern Recognition,Image and Video Processing

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the computer - aided detection of squamous cell carcinoma of the cervix (SCC). Specifically, the goal of the paper is to combine digital pathology and deep - learning techniques to improve the objectivity and efficiency of the analysis of squamous cell carcinoma of the cervix tissue sections, reduce the diagnostic variability caused by differences in pathologists' experience, and thus assist pathologists in accelerating the patient's diagnosis process. The paper mentions that traditional diagnosis of squamous cell carcinoma of the cervix mainly depends on histopathological evaluation by pathologists under a microscope. This method is highly dependent on the pathologists' experience, so there are large inter - observer and intra - observer variations. To overcome this challenge, the research team collected 800 hematoxylin and eosin (H&E) - stained sections from 300 patients with squamous cell carcinoma of the cervix. They used information on morphological heterogeneity to establish deep - learning models based on popular convolutional neural network architectures (such as Inception - v3, InceptionResnet - v2 and Resnet50), and introduced random forests for feature extraction and slice - based classification. The research results show that the proposed model performs excellently in slice - based tumor discrimination, with an AUC score exceeding 0.94, but performs generally in lesion location identification in the entire slice image (FROC score exceeding 0.52), mainly due to the extreme complexity of the tumor tissue. In conclusion, this study is the first to propose a quantitative visual method for analyzing squamous cell carcinoma of the cervix sections, demonstrating a way to use new computational methods to assist pathologists and accelerate patient diagnosis.

Computer-aided Detection of Squamous Carcinoma of the Cervix in Whole Slide Images

Artificial Classification of Cervical Squamous Lesions in ThinPrep Cytologic Tests Using a Deep Convolutional Neural Network.

Quantitative analysis of abnormalities in gynecologic cytopathology with deep learning

Robust whole slide image analysis for cervical cancer screening using deep learning

Accurate and reproducible invasive breast cancer detection in whole-slide images: A Deep Learning approach for quantifying tumor extent

Artificial intelligence-assisted fast screening cervical high grade squamous intraepithelial lesion and squamous cell carcinoma diagnosis and treatment planning

Dual-path network with synergistic grouping loss and evidence driven risk stratification for whole slide cervical image analysis

Computer-Assisted Quantitative Framework for Whole Slide Cervical Image Grading Driven by Time Series Features

Interpretable Classification from Skin Cancer Histology Slides Using Deep Learning: A Retrospective Multicenter Study

COMPUTER AIDED DIAGNOSTIC SYSTEM FOR WHOLE SLIDE IMAGE OF LIQUID BASED CERVICAL CYTOLOGY SAMPLE CLASSIFICATION USING CONVOLUTIONAL NEURAL NETWORK

Large-scale cervical precancerous screening via AI-assisted cytology whole slide image analysis

A Survey on Deep Learning-based Cervical Cytology Screening: from Cell Identification to Whole Slide Image Analysis

A Survey for Cervical Cytopathology Image Analysis Using Deep Learning.

Pathologist-level interpretable whole-slide cancer diagnosis with deep learning

Colorectal Cancer Detection Based on Deep Learning

Cell comparative learning: A cervical cytopathology whole slide image classification method using normal and abnormal cells

A Computer-Aided Diagnosis System for Breast Pathology: A Deep Learning Approach with Model Interpretability from Pathological Perspective

Accuracy and Efficiency of Deep-Learning–Based Automation of Dual Stain Cytology in Cervical Cancer Screening

Cervical Cancer Metastasis and Recurrence Risk Prediction Based on Deep Convolutional Neural Network

Fast Whole Slide Image Analysis of Cervical Cancer Using Weak Annotation.

Detection of Cervical Cancer Cells in Whole Slide Images Using Deformable and Global Context Aware Faster RCNN-FPN